An investigation has been described in which an attempt has been made to measure the losses of dry matter in a number of regularly spaced layers of silage contained in a tower silo, the immediate object being to determine the average loss of nutrient matter in the entire mass of material. The work was continued over the period of two silage seasons. The more important findings are summarised below:

1. It is usually asserted by writers on ensilage that the attainment of such high temperatures as are necessary for the production of “sweet” silage necessarily involves an excessive loss of the dry matter of the crop as a consequence of the oxidation of carbohydrate. This statement has been shown to be erroneous both on theoretical grounds and on the grounds of actual measurements of the losses of dry matter entailed in the production of “sweet” silage. The amount of destruction of carbohydrate bears no significant relation to the temperature attained during preservation, and the factors of (a) juice drainage, (b) bacterial decomposition of carbohydrate, (c) partial spoiling by undesirable bacterial activity, as with “sour” silage, are of much greater significance in causing unduly large losses of carbohydrate.

2. The results obtained in this investigation with crops containing from 26·5 to 33·9 per cent. of dry matter show that “acid brown” silage can be made in the tower silo with an average loss of dry matter equal to 5–6 per cent. of that contained in the green crop. It is further concluded that “sweet” silage, if made under good conditions, can also be produced in the tower silo with an average dry matter loss of the same order. With “green fruity” silage, the average loss under proper conditions of ensilage in tower silos is of the order of 8–9 per cent. The evidence affords strong disproof of the statement that the ensilage of green crops cannot be accomplished without large losses of nutrient matter.